INDEX

Explanations

elements related to scaling, detecting, and describing features in data or images

oai_token-act-pair · gpt-4o-mini Triggered by @bot

sequences of decreasing sizes

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mtrl

-0.53

extAlignment

-0.44

styleType

-0.41

hun

-0.37

MethodManager

-0.36

 فلا

-0.36

хьтан

-0.36

UnitTesting

-0.36

gran

-0.35

canestro

-0.35

POSITIVE LOGITS

 Viana

0.57

########.

0.56

 Usaha

0.51

#+#

0.50

 dAtA

0.47

 AssemblyVersion

0.47

 sizes

0.47

 amores

0.47

تقاوى

0.46

 pravi

0.46

Activations Density 0.214%

elements related to scaling, detecting, and describing features in data or images

sequences of decreasing sizes

No Comments

No Known Activations